This thesis is part of the ANR GRADIENT project, led by Hamamache Kheddouci. The aim is to find complete and relevant models of proteins, taking into account surface, structural and semantic properties. Several models exist in the literature, but at present none of these methods takes all these properties into account at the same time. The aim of this comprehensive modeling is to capture all the properties characterizing the protein and, at the same time, to be able to visualize it from several points of view and to analyze it from both visual and algorithmic points of view.
The different models will be developed, compared and evaluated.